Opencode Configuration
Opencode
Official guide: https://opencode.ai/v2/docs
CLI Version
Linux & macOS
- Download cli version:
curl -fsSL https://opencode.ai/v2/install | bash
- open the application on the terminal with the command:
opencode
- Then exit from the application with the command \quit or by pressing Ctrl+C
- Open the configuration file
.config/opencode/opencode.jsoncin your file explorer and replace its contents with the new settings below.
{
"$schema": "https://opencode.ai/config.json",
"provider": {
"elettragateway": {
"name": "ElettraGateway",
"npm": "@ai-sdk/openai-compatible",
"options": {
"baseURL": "https://llm-api-gw.elettra.eu/v1",
"apiKey": "sk_ePkBU....."
},
"models": {
"top": { "name": "top", "modalities": { "input": ["text", "image"] } },
}
}
}
}
Windows
We suggest you to use the Windows Subsystem for Linux (WSL)
After activating it, use the same procedure as described in Linux & macOS to install and use it!
DESKTOP Version
-
Download the correct installer for your operating system from
https://opencode.ai/v2/docs#desktopand complete the installation. -
Edit the configuration file
- On Linux/macOS:
~/.config/opencode/opencode.jsonc - On Windows:
C:\Users\<YourUser>\.config\opencode\opencode.jsonc
Replace its contents with the settings you need (ensure the file is valid JSONC).
- On Linux/macOS:
-
Launch Opencode from a terminal or by double‑clicking the application icon.
-
In Opencode, go to File → Settings → Models.
- Enable the top model.
- Disable any other models that are listed.
-
In the prompt bar, select the model you prefer (the one you just enabled).
You’re now ready to use it.